Algorithm Based on Multiple Features for Recognition of E.coli Promoter Region
A promoter plays an important role in the transcription of DNA sequences; its recognition is very significant. This paper concerns about the problem of extracting information from a given DNA region and to make predictions whether it contains a promoter. We incorporate three types of information to access the significance of a promoter, information on content and information on structure, and then these information are used in a trained nonlinear classifier. The results show that our method performs well and achieve 13.4% average error rate on positive and negative data from non-coding regions and 17.0% average error rate on positive and negative data from coding regions.
W. Ao et al., "Algorithm Based on Multiple Features for Recognition of E.coli Promoter Region", Applied Mechanics and Materials, Vols. 48-49, pp. 164-169, 2011